# Gzip\_disable doesn't disable gzip\_vary **URL:** \ **Category:** NGINX\ **Created:** [March 2, 2012, 8:09am UTC](https://www.ruby-forum.com/t/gzip-disable-doesnt-disable-gzip-vary/217467 "2012-03-02T08:09:07Z")\ **Posts on this page:** 6\ **Page:** 1
**Author:** ![dubstep](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/d/cdc98d/32.png) [@dubstep](https://www.ruby-forum.com/u/dubstep)\ **Post date:** [March 2, 2012, 8:09am UTC](https://www.ruby-forum.com/t/gzip-disable-doesnt-disable-gzip-vary/217467/1 "2012-03-02T08:09:07Z")
I noticed that with this setting: gzip on; gzip\_vary on; gzip\_disable “msie6”; Header “Vary: Accept-Encoding” will always be returned even though the user agent is ie6 (and therefore not using gzip): # curl -I [http://animebsd.net/about/feed](http://animebsd.net/about/feed) -A "Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.0)" HTTP/1.1 200 OK Server: nginx/1.1.14 Date: Fri, 02 Mar 2012 07:04:25 GMT Content-Type: text/xml; charset=UTF-8 Connection: keep-alive Vary: Accept-Encoding X-Pingback: [http://animebsd.net/xmlrpc.php](http://animebsd.net/xmlrpc.php) Last-Modified: Fri, 02 Mar 2012 04:17:00 GMT ETag: “92a5989a0949b85e303228490256ebf2” X-Powered-By: W3 Total Cache/0.9.2.4 Link: [http://wp.me/Pig2n-5B](http://wp.me/Pig2n-5B); rel=shortlink I don’t think this is intentional as getting non-gzipped content doesn’t return such header. (I’ve verified this is returned by nginx by running same test with disabled gzip\_vary) (I don’t see mention of this being fixed in .15 and .16 in the changelog so I haven’t tested with them) ---
**Author:** ![Edho\_A](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/e/848f3c/32.png) [@Edho\_A](https://www.ruby-forum.com/u/Edho_A)\ **Post date:** [March 2, 2012, 10:59am UTC](https://www.ruby-forum.com/t/gzip-disable-doesnt-disable-gzip-vary/217467/2 "2012-03-02T10:59:55Z")
Hello! On Fri, Mar 02, 2012 at 02:08:05PM +0700, Edho A. wrote: > MSIE 6.0; Windows NT 5.0)" > Link: [http://wp.me/Pig2n-5B](http://wp.me/Pig2n-5B); rel=shortlink > > I don’t think this is intentional as getting non-gzipped content > doesn’t return such header. (I’ve verified this is returned by nginx > by running same test with disabled gzip\_vary) This is intentional, “Vary” is returned whenever gzip (and gzip\_vary) is enabled, as response representation depends on header(s) listed. Strictly speaking, it should also include User-Agent in case of gzip\_disable defined. Maxim D. ---
**Author:** ![Edho\_A](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/e/848f3c/32.png) [@Edho\_A](https://www.ruby-forum.com/u/Edho_A)\ **Post date:** [March 2, 2012, 11:16am UTC](https://www.ruby-forum.com/t/gzip-disable-doesnt-disable-gzip-vary/217467/3 "2012-03-02T11:16:03Z")
2012/3/2 Maxim D. [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid): > This is intentional, “Vary” is returned whenever gzip (and > gzip\_vary) is enabled, as response representation depends on > header(s) listed. > > Strictly speaking, it should also include User-Agent in case of > gzip\_disable defined. Shouldn’t there be no change to header at all when client user agent matches the gzip\_disabled (ie. just like with gzip off)? ---
**Author:** ![Edho\_A](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/e/848f3c/32.png) [@Edho\_A](https://www.ruby-forum.com/u/Edho_A)\ **Post date:** [March 2, 2012, 12:21pm UTC](https://www.ruby-forum.com/t/gzip-disable-doesnt-disable-gzip-vary/217467/4 "2012-03-02T12:21:17Z")
Hello! On Fri, Mar 02, 2012 at 05:15:07PM +0700, Edho A. wrote: > Shouldn’t there be no change to header at all when client user agent > matches the gzip\_disabled (ie. just like with gzip off)? > **[RFC 2616: Hypertext Transfer Protocol -- HTTP/1.1 | RFC Editor](https://www.rfc-editor.org/info/rfc2616/)** > > HTTP has been in use by the World-Wide Web global information initiative since 1990. This specification defines the protocol referred to as "HTTP/1.1", and is an update to RFC 2068. \[STANDARDS-TRACK\] An HTTP/1.1 server SHOULD include a Vary header field with any cacheable response that is subject to server-driven negotiation. For sure with gzip enabled the response “is subject to server-driven negotiation”. Maxim D. ---
**Author:** ![Edho\_A](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/e/848f3c/32.png) [@Edho\_A](https://www.ruby-forum.com/u/Edho_A)\ **Post date:** [March 2, 2012, 12:43pm UTC](https://www.ruby-forum.com/t/gzip-disable-doesnt-disable-gzip-vary/217467/5 "2012-03-02T12:43:00Z")
2012/3/2 Maxim D. [removed\_email\_address@domain.invalid](mailto:removed_email_address@domain.invalid): > So, the “gzip enabled response” includes cases with non-gzip content as: - no “Accept-Encoding: gzip,deflate” - matches gzip\_disabled and therefore “Vary: Accept-Encoding” will always be added even though the response isn’t compressed? Or in other words, does “gzip on; gzip\_vary on” means “add ‘Vary: Accept-Encoding’ header on any responses matching gzip\_types regardless whether the response actually compressed or not”? ---
**Author:** ![Edho\_A](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/e/848f3c/32.png) [@Edho\_A](https://www.ruby-forum.com/u/Edho_A)\ **Post date:** [March 2, 2012, 1:33pm UTC](https://www.ruby-forum.com/t/gzip-disable-doesnt-disable-gzip-vary/217467/6 "2012-03-02T13:33:55Z")
Hello! On Fri, Mar 02, 2012 at 06:41:50PM +0700, Edho A. wrote: > > negotiation". > > regardless whether the response actually compressed or not"? > > Yes. Maxim D.